During development, cell fate specification depends on both reproducible and stochastic regulatory mechanisms. Very little is known about stochastic mechanisms that turn gene expression on or off randomly in individual cells to diversify fates. Random cell fate choices are important for the diversification of visual and olfactory receptors, specification of motor neuron subtypes, dendritic self-avoidance in neurons, determination of immune cell fates, and differentiation of stem cells. Studying stochastic gene expression will enhance our understanding of vision disorders, anosmia, autism, immunodeficiencies, and lymphoma. The main goal of our work is to address how gene expression is regulated in a stochastic on/off manner during development using specification of Drosophila color photoreceptors as a model. The color vision system of the fly is a random mosaic of two photoreceptor subtypes determined by the stochastic on/off expression of the transcription factor Spineless. We recently discovered that each copy of the spineless gene makes an intrinsic, random expression decision within a single nucleus exposed to the same milieu of trans-acting factors. Our studies suggest that stochastic spineless expression requires regulation by specific transcription factors in R7 photoreceptors, insulator DNA elements that mediate DNA looping and Polycomb Response Elements/Trithorax Response Elements (PREs/TREs) that regulate expression. We hypothesize that a unique combination of transcription factors expressed in all R7s is required for stochastic ss expression (Aim 1). We propose that stochastic on/off expression is dictated by the ss locus randomly assuming one of two DNA looping configurations mediated by insulators (Aim 2). These configurations determine the proximity of a repressive PRE/TRE to the ss promoter. In on cells, the repressive PRE/TRE remains distant from the promoter to allow activation; in off cells, the repressive PRE/TRE is proximal to the promoter to repress expression (Aim 3). We will test these hypotheses by assessing transcription factor binding and combinatoriality, determining functionality of insulators and PRE/TRE DNA elements by generating deletions using CRISPR/Cas9 and conducting transgene assays, and examining the DNA looping conditions and histone modifications in the Spineless on and Spineless off cells using the recently developed cgChIP technology. Achieving these aims will characterize how regulation by trans factors, insulator-mediated DNA looping, and chromatin state control stochastic gene expression. Our findings will have a profound impact on our understanding of how the 3D architecture of the nucleus and epigenetic modifications control gene expression.